By Martin Baccardax
Oracle shares have lost nearly $400 billion in value over the past four months, to near their lowest levels since early June. They are seen as a market-listed proxy for both OpenAI and the titanic amounts of capital being committed to the artificial intelligence investment boom.
Guggenheim analysts, however, led by John DiFucci, think the cloud and software company is a "decade stock" that will see growth at a "hyper rate" and cash flows that will "waterfall" over the longer term.
"We continue to view Oracle as our Best Idea and a stock that could work over the next decade," DiFucci and his team said in a note published Tuesday. "We believe that it simply has a better mousetrap in cloud infrastructure built on lessons learned by others that came before it (primarily Amazon Web Services), coupled with unique enterprise grade technologies that enable scaling and security like no others can today."
The bank now carries a $400 price target on Oracle stock, a level that would imply a 118% vault from where it trades currently. Oracle stock was down around 4.2% on Tuesday, and changing hands at $183.10, amid a broad tech-led selloff. The Nasdaq was down around 1.6%.
At the core of its thesis, DiFucci thinks Oracle would need to spend around $290 billion on AI infrastructure over the next five years, and would need to borrow around $112 billion by the end of its 2028 fiscal year to meet its goal of generating $166 billion in annual revenue by 2030.
Adding $112 billion in new borrowing would take Oracle's total debt levels past $200 billion, but the bank isn't sure that's the base case.
Not all of that cash would need to come directly from the bond market, DiFucci and his team argue, and that would allow the group to keep both its investment grade credit rating and maintain a lower cost of capital.
In fact, the bank sees Oracle possibly issuing new stock, in much the same way that Boeing did in 2024, to "signal to debt investors that Oracle is very serious about maintaining its investment grade debt rating."
"Since chairman Larry Ellison owns 41% of the equity of Oracle, this decision would largely rest with him," DiFucci and his team said. "We believe he's a very logical person, and if he believed that if Oracle sold equity, it would increase the value of his remaining equity by more than if the company didn't."
Oracle could also lower its overall capex further by arranging so-called "bring your own chip" deals with potential clients, which would see new customers providing their own AI GPUs to a new Oracle Cloud Infrastructure ( OCI) project. Guggenheim analysts, in fact, think those GPUs account for between 45% and 65% of a project's capex.
That would help Oracle boost margins and profits while lowering borrowing costs and interest expense. Earnings and cash flow would improve, as well.
A lingering issue for investors, however, remains Oracle's close financial ties to OpenAI, and its reliance on the ChatGPT creator for billions of dollars in future revenue.
Including Oracle, OpenAI has made around $1.4 trillion in spending commitments against current annual revenue of around $13 billion.
But DiFucci and his team see it differently, even as they note that OpenAI "doesn't currently have the obvious financial foundation to fulfill its financial commitments." They also think the $1.4 trillion figure is overstated.
"We believe with this risk comes significant opportunity," the bank said. "OpenAI will remain as at least one of the most important AI companies despite recent strides by others (including Google's recent win of Gemini to power Apple's Siri assistant) and will remain a significant OCI customer".
It's also getting support from Microsoft, which owns a 27% stake in its for-profit division valued at $135 billion, Advanced Micro Devices, which cut a chips-for-equity deal last autumn, and a letter of intent that could see Nvidia investing as much as $100 billion.
Collectively, Oracle's likely improving debt position, its market-leading technology, and the potential for billions in revenue from OpenAI and others make it a compelling investment, according to DiFucci and his team.
But it still might take time to come good.
"We see this as an opportunity for investors with a longer time horizon, a few years rather than a few weeks or months, to start to build a position," he said. "While we keep calling this a decade stock (because we think it is), we expect the numbers to be irrefutable in half that time and be obvious long before that."
Write to Martin Baccardax at martin.baccardax@barrons.com
This content was created by Barron's, which is operated by Dow Jones & Co. Barron's is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
January 20, 2026 12:58 ET (17:58 GMT)
Copyright (c) 2026 Dow Jones & Company, Inc.
Comments